Negative 2D points in image with pinhole transform lock 2D view to top-left corner

[roym899]

Describe the bug
Logging negative 2D points in a 2D image will make it impossible to pan the 2D view of the image (it will be locked to the bottom left corner).

To Reproduce

import numpy as np
import rerun as rr

rr.init("images", spawn=True)

rr.log_image(
    f"world/camera/image/rgb", np.random.rand(500, 500, 3)
)
rr.log_pinhole(
    f"world/camera/image",
    child_from_parent=[[500, 0, 250], [0, 500, 250], [0, 0, 1]],
    width=500,
    height=500,
)
rr.log_points(
    f"world/camera/image/points", [[100,-100], [0, 0], [100, 100]]
)

And zoom into the image.

Remove the first point and things work fine. Remove the pinhole and things work fine as well.

Desktop (please complete the following information):

• OS: Ubuntu 20.04

Rerun version

rerun_py 0.7.0 [rustc 1.69.0 (84c898d65 2023-04-16), LLVM 15.0.7] x86_64-unknown-linux-gnu prepare-0.7 9cf3033, built 2023-06-16T15:47:32Z

[emilk]

Updated version:

import numpy as np
import rerun as rr

rr.init("images", spawn=True)

rr.log("world/camera/image/rgb", rr.Image(np.random.rand(500, 500, 3)))
rr.log(
    "world/camera/image", rr.Pinhole(image_from_camera=[[500, 0, 250], [0, 500, 250], [0, 0, 1]], width=500, height=500)
)
rr.log("world/camera/image/points", rr.Points2D([[100, -100], [0, 0], [100, 100]]))